Identification of the change point in panel data using simultaneously EWMAA and CUSUM

Identification of the change point in panel data leads practitioners to focus on the time when really a change takes place in the cross sectional data. Identification of the time helps one to provide a more realistic analysis of the change manifested itself to the process. Different methods of change point identification have been proposed in literature, however, the literature addressees that the sensitivity of identifying the change point is an important issue. This paper attempts to propose a new method with high sensitivity for identifying the change point in a panel data (with large dimension) through a hybrid approach. The proposed method is named Double CUSUM-EWMA. The comparative report addresses that the performance of the proposed method has relatively better performance compared to the existing methods in the literature. This study analyzes several simulated numerical examples with large dimension of panel data when a step shift manifests itself to the process.